Environment: What makes fine dust so dangerous

Inflammation, bronchitis, asthma attacks, cardiovascular problems ... - an excerpt from the list of possible health problems caused by high concentrations of particulate matter: particles with a diameter of ten micrometers or less - called "PM10" - and even smaller "PM2.5" particles that enter our lungs from car exhaust, heating systems, industrial operations and natural sources. Although strict monitoring and abatement measures have reduced pollution levels in Switzerland since the 1990s, the problem persists in many places, especially in cities.

Particle size, composition, sources and effects of particulate matter are not easy to determine. One thing is certain: the smaller the particles, the deeper they reach the human lungs. But which fractions are particularly dangerous? And in what combinations and concentrations? To describe this, the new criterion "oxidative potential" was developed a few years ago: a term intended to describe the ability of inhaled particles to trigger the formation of so-called free radicals in the body, which can ultimately lead to inflammation.

Measurement campaign at five Swiss locations

Empa researchers Stuart Grange and Christoph Hüglin from the "Air Pollutants / Environmental Technology" department have investigated the suitability of this measure of health risk for Switzerland in greater detail - with an elaborate measurement campaign supported by the French "Université Grenoble Alpes". With the help of the National Air Pollutant Monitoring Network (NABEL), which Empa operates with the Federal Office for the Environment (FOEN), they collected particulate matter samples in the PM2.5 and PM10 categories around the clock between June 2018 and May 2019. The measuring stations covered the entire range of particulate pollution and were located in cities, the agglomeration and in rural areas south and north of the Alps.

In total, the experts analyzed around 900 samples in the laboratory - using test methods for oxidative potential that work with different analysis substances: Ascorbic acid (AA for short), dithiothreitol (DTT) and dichlorofluorescein (DFCH). In the AA test, the consumption of ascorbic acid, an important antioxidant, allows conclusions to be drawn about the oxidative "toxicity" of the particulate matter sample, for example due to contained metals. The other two methods work in a similar way, but use different substances for detection. Put simply, says Christoph Hüglin, the three methods offer different perspectives on similar biological processes.

Machine learning reduces complexity

In addition to the oxidative potential, a large number of chemical constituents of fine dust were analyzed. This resulted in a large amount of data on the elements, ions and organic substances that make up the fine dust in Switzerland. In order to identify the "suspicious substances" with the greatest oxidative potential in these masses of data, the Empa researchers used "machine learning" methods. More precisely, they used the "Random Forest" algorithm, which, metaphorically speaking, allows a forest of countless trees to grow, each of which makes decisions about relationships in the data - such as, in this case, the ingredients of particulate matter and the associated oxidative potential. In the end, a mean model is formed from the decisions of the entire forest.

In this way, the experts reduced the number of suspect ingredients to around a dozen, which they in turn analyzed using conventional calculation methods and models to finally track down the most important health hazards - various metals or even organic substances, which in turn provide clues to their origin and causes.

The results confirm known facts such as a clear urban-rural divide in particulate matter and its health consequences, as well as higher pollution in winter than in summer. Admittedly, there were exceptions: The values for the oxidative potential, related to the air volume, had increased particularly significantly in the south of Switzerland during the cold season - in areas that were polluted by smoke from wood combustion during this time.

The lowest mean values were found in rural areas, while the highest values over the entire period came from an urban and traffic-heavy measuring station. At busy intersections in cities, other emissions besides exhaust gases are a cause for concern: metals such as copper, zinc, and manganese indicate particulate matter components that may originate from the abrasion of car tires or brake pads, for example.

How exactly the criterion of oxidative potential can describe health hazards is currently the subject of controversial debate among experts. After all, even the most precise measurements and analyses of air pollutants do not answer open questions about inflammatory processes in the human body. But Empa researcher Hüglin, after the analyses with his team, at least assumes that meaningful measures can be derived from them: Although all fine dust particles could affect health - but with regard to the oxidative potential, the components from road traffic, which do not originate from exhaust gases, and from wood combustion should be given special attention in measures to protect the population.

Dangers due to fine dust

Particulate matter can be described as a mixture of solid and also liquid particles in the air - from anthropogenic sources such as engines or industrial exhaust gases or also natural sources such as volcanoes. While many particles enter the air directly through emissions (primary particles), secondary particles are first formed in the atmosphere through chemical reactions of gaseous compounds. Of particular importance for human health is respirable fine dust, whose particles have an aerodynamic diameter of less than ten micrometers. Particulate matter also includes so-called ultrafine particles, such as those from diesel engine exhaust, which penetrate deep into the lungs and can cause severe damage.

Source: Empa